The invention concerns a device and a method for providing information about animals when walking through an animal passage, whereby this information includes at least the number of animals walking through the animal passage.
A generic method and a generic device are known from DE 103 51 911 A1. The method proposed in this publication for recognition of the passage of animals when walking through an animal passage as entry to an automatic milking parlor comprises first an identification of the animals to be milked using an identification device. For example, this identification device is designed according to the manner of an RFID system, which comprises an emitter/receiver unit on a passage door or railing and a transponder or responder (both RFID tags) optionally with sensor, carried by the animals to be detected.
The electronic identification of animals using only such identification devices may result in detection problems in certain operational situations. Normally the animals enter the milking parlor through an entry area. This is where the identification device is located which serves for the individual identification of the animals. In this way, in a group milking parlor, an assignment between the animals to be milked and the milking stations can be carried out. Thus, the first animal that was recognized by the identification device is automatically assigned to the first milking station, the second animal to the second milking station, and so on.
However, if an animal has lost its detection mark or its responder or if this is defective, or if the detection is perturbed, it is not noted that an animal already physically occupies a milking station, and this appears to be free to the computer device which controls and manages the assignment of the milking station on the basis of a herd-management system or program.
As a result of this, for all subsequent animals of the series, there will be a shift between animal and milking station. The animals are in that case still milked, but the amounts of milk and the other data obtained under certain circumstances (milking duration, conductance, etc.), are discarded, in order to avoid the storage of incorrect data in the herd-management system.
In order to avoid this problem, in the generic DE 103 51 911 A it is therefore further proposed that the animals be counted independently of the identification of their responder, using a detection device which—based on a sensor system such as a photoelectric barrier or a camera—captures and counts the animals.
In this way it is achieved that a non-identified animal that stands at a milking station will nevertheless be included in the milking station occupation record so at least the subsequent identified animals will automatically be assigned to the correct milking stations.
Thus, the generic device has proven to be useful. However, it seems that it can still be optimized from the point of view of the design of the detection device. It turns out that exact counting of the animals proves to be difficult since unlike in the usual industrial processes they do not enter the milking parlor at regular distances from one another or at a regular velocity. Furthermore, it can occur that an animal remains standing in the detection area so that two animals enter the milking parlor placed directly one behind the other (FIG. 6) or sideways behind one another (FIG. 7) or that individual or several animals run into the milking parlor at high speed (indicated in FIG. 5). All these situations are difficult to capture.
Regarding the technological background, let us recall the following state of the art:
WO 03/039247 A1 concerns an installation for the control of the presence of an identification label (responder) with animals. An animal that cannot be identified when the identification label is missing should still be recognized in a reliable manner.
Additional devices must be used for this purpose: a detection device which determines whether the animal can be found in the animal recognition area and an additional discriminating device which marks the animal when it has no responder and correspondingly no signal was detected in the animal recognition area. This discriminating device can mark, for example, by spraying the animal that has lost its responder with paint, which is relatively expensive.
EP 1 537 531 B1 describes an image recording system and an image recording method for evaluation of the physical condition of animals. Hereby the body condition score (the numerical value BCS) is determined, whereby the method comprises the following three steps:                delineating a predetermined section of interest on the animal body and producing data that represent this;        processing the produced data in order to obtain a three-dimensional image of the section of interest and        analyzing the three-dimensional design, in order to determine a predefined measurable parameter, whose surface relief represents the section of interest in order to indicate the state of the body.        
WO 03/059191 concerns a method for reliable prediction of the body condition score (numerical value) for animals. The determined numerical value for the body condition score (BCS) can be used to determine the correct food rations for domestic animals.
WO 2004/049790 concerns a device for the detection of an animal with a body part and a head part, whereby parameters are determined that refer to the size of the cross section of the animal at a specific position in the passage.
Based on this background, starting from the generic description, the invention has first of all the task to further develop the generic device and the generic method in such a way that the number of erroneous assignments during the occupation of the milking station is minimized and as many data as possible are obtained about the animal.
With regard to the method according to the invention for providing information about animals when they walk through an animal passage with a device according to the invention, with the detection device, animal data of the animals walking through the animal passage are captured and evaluated with the evaluation device and counting pulses are produced based on this evaluation in order to set an animal counter of the evaluation device or processor, whereby the detection of the animals comprises at least the steps of producing 3D data from 3D images of the animals or of partial areas of the animals and an evaluation of these 3D data to produce at least the counter pulse.
Furthermore, preferably the method comprises identification of the animals with the aid of an identification device, optionally directly from data obtained by the detection device.
Furthermore, preferably the method includes determination of another animal parameter from the data obtained by the detection device, whereby the additional animal parameter can be a parameter that characterizes the state of the health of the animal, which is determined from the 3D images.
It is expedient for the additional animal parameters to be continuously stored in a database during each milking process.
The device increases the detection rate during passage recognition in a simple manner. In addition, with the device, detection at individual stations can be omitted in large milking parlors.